Statistical learning theory 2023/24

Материал из Wiki - Факультет компьютерных наук
Версия от 16:02, 13 сентября 2024; Bauwens (обсуждение | вклад)

(разн.) ← Предыдущая | Текущая версия (разн.) | Следующая → (разн.)
Перейти к: навигация, поиск

General Information

Lectures: Tuesday 14h40 -- 16h00, room S321 and in zoom by Bruno Bauwens and Maxim Kaledin,

Seminars: Monday 16h20 -- 17h40, room N506, and in zoom by Artur Goldman.

The course is similar to last year.


Colloquium

Rules and questions.

Date: Tuesday December 19th during the lecture. (It is possible to come on 12.12 during the lecture or on 12.19 after 18h10 to room ??, but notify Bruno by email.)


Problems exam

December 22th, 13h--16h, room D507, (it is a computer room).
-- You may use handwritten notes, lecture materials from this wiki (either printed or through your PC), Mohri's book
-- You may not search on the internet or interact with other humans (e.g. by phone, forums, etc)


Course materials

Video Summary Slides Lecture notes Problem list Solutions
Part 1. Online learning
05 Sept Philosophy. The online mistake bound model. The halving and weighted majority algorithms. sl01 ch00 ch01 prob01 sol01
12 Sept The perceptron algorithm. Kernels. The standard optimal algorithm. sl02 ch02 ch03 prob02 sol02
19 Sept Prediction with expert advice. Recap probability theory (seminar). sl03 ch04 ch05 prob03 sol03
26 Sept Multi-armed bandids. notes04 prob04
Part 2. Distribution independent risk bounds
03 Oct Necessity of a hypothesis class. Sample complexity in the realizable setting, examples: threshold functions and finite classes. sl04 ch06 prob05 sol05
10 Oct Growth functions, VC-dimension and the characterization of sample comlexity with VC-dimensions sl05 ch07 ch08 prob06 sol06
17 Oct Risk decomposition and the fundamental theorem of statistical learning theory sl06 ch09 prob07 sol07
24 Oct Bounded differences inequality, Rademacher complexity, symmetrization, contraction lemma. sl07 ch10 ch11 prob08 sol08
Part 3. Margin risk bounds with applications
07 Nov Simple regression, support vector machines, margin risk bounds, and neural nets with dropout regularization sl08 ch12 ch13 prob09 sol09
14 Nov Kernels: RKHS, representer theorem, risk bounds sl09 ch14 prob10 sol10
21 Nov AdaBoost and the margin hypothesis sl10 ch15 prob11 sol11
28 Nov Implicit regularization of stochastic gradient descent in overparameterized neural nets (recording with many details about the Hessian) ch16 ch17
05 Dec Part 2 of previous lecture: Hessian control and stability of the NTK.
12 Dec Colloquium (you may choose between 12 Dec and 19 Dec).

Background on multi-armed bandits: A. Slivkins, [Introduction to multi-armed bandits https://arxiv.org/pdf/1904.07272.pdf], 2022.

The lectures in October and November are based on the book: Foundations of machine learning 2nd ed, Mehryar Mohri, Afshin Rostamizadeh, and Ameet Talwalker, 2018. This book can be downloaded from Library Genesis (the link changes sometimes and sometimes vpn is needed).


Grading formula

Final grade = 0.35 * [score of homeworks] + 0.35 * [score of colloquium] + 0.3 * [score on the exam] + bonus from quizzes.

All homework questions have the same weight. Each solved extra homework task increases the score of the final exam by 1 point.

There is no rounding except on the final grade. Arithmetic rounding is used.

Autogrades: if you only need 6/10 on the exam to pass with maximal final score, it will be given automatically. This may happen because of extra questions and bonuses from quizzes.


Homeworks

Deadline every 2 weeks, before the seminar at 16h00. Homework problems from

seminars 1 and 2 on September 25, seminars 3 and 4 on October 9, seminars 5 and 6 on November 6, seminars 7 and 8 on November 13, seminars 9 and 10 on November 27 December 4, seminar 11 before the start of the exam.

Email to brbauwens-at-gmail.com. Start the subject line with SLT-HW.

Late policy: 1 homework can be submitted at most 24 late without explanations. 3 HW tasks that were not submitted before can be submitted at any moment before the beginning of the exam.


Office hours

Bruno Bauwens: Wednesday 13h-16h, Friday 14h-20h, (better send an email in advance).

Maxim Kaledin: Write in Telegram, the time is flexible

Artur Goldman: Write in Telegram, the time is flexible